Exploring Models and Data for Remote Sensing Image Caption Generation

By Xiaoqiang Lu, Binqiang Wang, Xiangtao Zheng, Xuelong Li. IEEE Transactions on Geoscience and Remote Sensing, 2017

709 citations recorded by the source corpus.

Inspired by recent development of artificial satellite, remote sensing images have attracted extensive attention. Recently, notable progress has been made in scene classification and target detection. However, it is still not clear how to describe the remote sensing image content with accurate and concise sentences. In this paper, we investigate to describe the remote sensing images with accurate and flexible sentences. First, some annotated instructions are presented to better describe the remote sensing images considering the special characteristics of remote sensing images. Second, in order to exhaustively exploit the contents of remote sensing images, a large-scale aerial image data set is constructed for remote sensing image caption. Finally, a comprehensive review is presented on the proposed data set to fully advance the task of remote sensing caption. Extensive experiments on the proposed data set demonstrate that the content of the remote sensing image can be completely described by generating language descriptions. The data set is available at https://github.com/201528014227051/RSICD_optimal.

View Paper: https://www.semanticscholar.org/paper/7cc21bcce77dacd1ea7646244043261167f2dcd0

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

RSGPT: A Remote Sensing Vision Language Model and Benchmark (2023, unknown): “…which can provide valuable reference for user interaction of remote sensing knowledge through VLMs. 2) Qualitative Analysis: We first compare RSICap with previous RS image caption datasets in Figure 1, including UCM-Captions [44], Sydney-Captions [44], RSICD [36], NWPU-Captions [10] and RS5M [79].”
Transformers in Remote Sensing: A Survey (2022, methodology): “TRL [103] Image Captioning RSICD [104], UCM-captions [105], Sydney-Caption [106] BLEU,”
EarthGPT: A Universal Multimodal Large Language Model for Multisensor Image Comprehension in Remote Sensing Domain (2024, methodology): “The NaSC-TG2 dataset consists of 20 000 RS images of 10 land cover types.”
RemoteCLIP: A Vision Language Foundation Model for Remote Sensing (2023, methodology): “This dataset is 12 times larger than the combination of RSITMD, RSICD, and UCM datasets [28], [29], [30], addressing the scarcity of pretraining data in remote sensing.”